Synaptic pruning in the central nervous system is a critical stage for development of a precise and functional neural network. While activity-dependent competition between presynaptic inputs has previously been demonstrated, it is unknown how competing synaptic inputs translate different activity levels into molecular pathways resulting in physical elimination of less active inputs. We hypothesize that the canonical effector molecule of apoptosis, caspase-3 (casp3), may be selectively activated at less active synapses during developmental pruning. We propose that casp3 activity could lead to the exposure of the membrane phospholipid phosphatidylserine (PS) to the extracellular mileu, allowing complement proteins to be recruited to the synapse and flag it for elimination. To study this pathway, we are using the mouse retinogeniculate system, a classic model of activity-dependent central nervous system (CNS) synapse elimination. We have developed both in vitro and in vivo models of synaptic competition and synapse elimination. Using these models, we will perform a variety of fixed and live imaging experiments to determine whether this apoptotic cascade is involved in developmental pruning of inappropriate synapses. Our findings can help shed light on the mechanisms of neurodevelopmental disorders thought to involve aberrant synaptic connectivity, such as autism spectrum disorders (ASDs).